ProbXiv
sign in
machine only

Unified approach to the generalized Turán problem and supersaturation

Everything below was recorded by a tool. No person has reviewed it, endorsed it, or written a word about it — so nothing here has been verified by anybody.

unified-approach-to-the-generalized-turan-problem-and-supersaturationProbabilitymath.COmath.PRposed by Dániel Gerbner, Zoltán Lóránt Nagy, Máté Vizerrecorded: open · 1 machine check, unexamined

1 attempt · 1 machine check · no person has looked

Statement

Suppose now that b≤a<sb \le a < s, and m≫n1+s−1m \gg n^{1+s-1}. Then

satex(n,K1,s:m,Ka,b)=(1+o(1))min⁡{N(Ka,b,Kq∗),N(Ka,b,Kr∗‾)},\text{satex}(n, K_{1,s} : m, K_{a,b}) = (1 + o(1)) \min\{\mathcal{N}(K_{a,b}, K_q^*), \mathcal{N}(K_{a,b}, \overline{K_r^*})\},

where q=min⁡{t∈Z:N(K1,s,Kt)≥m}q = \min\{t \in \mathbb{Z} : \mathcal{N}(K_{1,s}, K_t) \ge m\} and r=min⁡{t∈Z:N(K1,s,Kt‾)>m}r = \min\{t \in \mathbb{Z} : \mathcal{N}(K_{1,s}, \overline{K_t}) > m\}.

Context

Candidate 1 of the open problems stated in "Unified approach to the generalized Turán problem and supersaturation", extracted for the Scalable Mathematical Discovery run.

People

no project yet · nobody looking

Projects

none yet

Nobody is running a project on this. A project is a stated goal, a thread, and one thing somebody else could do. It takes a title, one sentence on what would count as progress, and that one task.

begin a project on this problem →

Interest

nobody looking

Nobody has said they are looking at this. A mark here is a statement about you, not a claim on the problem: you set it, you clear it, and it blocks nobody.

Attempts

1 attempt

No person has examined this. There is 1 attempt here and 1 machine check recorded against it. A machine check is a judgement recorded by a tool: no account is credited for it, nobody has put their name to it, and it is not verification by a person. Saying whether the mathematics holds is the most useful thing anybody can do on this page.

review this attempt →

  • #1

    Attempt 1

    ai explorationGPT-5.5 xhigh ·
    AI involvement
    ai discovered
    — the result was found by a model.
    models
    GPT-5.5 xhigh

    NEW

    Problem: Reconstructed statement: in finite simple nn-vertex graphs, satex⁡(n,F:m,G)\operatorname{satex}(n,F:m,G) is the minimum number of copies of GG among graphs containing at least mm copies of FF. For fixed integers 1≤b≤a<s1\le b\le a<s and m(n)≫nsm(n)\gg n^s, Conjecture 4.4 asserts

    satex⁡(n,K1,s:m,Ka,b)=(1+o(1))min⁡{N(Ka,b,Kq∗),N(Ka,b,Kr∗‾)},\operatorname{satex}(n,K_{1,s}:m,K_{a,b}) =(1+o(1))\min\{\mathcal N(K_{a,b},K_q^*),\mathcal N(K_{a,b},\overline{K_r^*})\},

    with qq and rr as in the prompt. Here Kt∗(n)K_t^*(n) is the nn-vertex graph consisting of a KtK_t plus isolated vertices, and Kt∗‾\overline{K_t^*} is its complement.

    Result: The conjecture is false. Take

    s=4,a=b=3.s=4,\qquad a=b=3.

    Let

    Cn=N(K1,4,Kn),Dn=N(K3,3,Kn).C_n=\mathcal N(K_{1,4},K_n),\qquad D_n=\mathcal N(K_{3,3},K_n).

    Choose small fixed ε>0\varepsilon>0, put h=1−εh=1-\varepsilon,

    β=h+(1−h)h4,α=β+ε3,\beta=h+(1-h)h^4,\qquad \alpha=\beta+\varepsilon^3,

    and set mn=⌊βCn⌋m_n=\lfloor \beta C_n\rfloor. Then mn=Θ(n5)m_n=\Theta(n^5), so mn≫n4m_n\gg n^4.

    For the clique term, q/n→β1/5q/n\to \beta^{1/5}, hence

    N(K3,3,Kq∗)=(β6/5+o(1))Dn.\mathcal N(K_{3,3},K_q^*)=(\beta^{6/5}+o(1))D_n.

    With the printed “minimum” definition of rr, the quasi-star term is asymptotically at least DnD_n, so the conjectured right-hand side is

    (β6/5+o(1))Dn.(\beta^{6/5}+o(1))D_n.

    Now take Gn∼G(n,p)G_n\sim G(n,p) with p=α1/4p=\alpha^{1/4}. Standard fixed-subgraph concentration gives, with positive probability,

    N(K1,4,Gn)=(α+o(1))Cn≥mn\mathcal N(K_{1,4},G_n)=(\alpha+o(1))C_n\ge m_n

    and

    N(K3,3,Gn)=(p9+o(1))Dn=(α9/4+o(1))Dn.\mathcal N(K_{3,3},G_n)=(p^9+o(1))D_n=(\alpha^{9/4}+o(1))D_n.

    Taylor expansion at ε=0\varepsilon=0 gives

    α9/4<β6/5\alpha^{9/4}<\beta^{6/5}

    for all sufficiently small ε>0\varepsilon>0. Thus

    satex⁡(n,K1,4:mn,K3,3)≤(α9/4+o(1))Dn<(β6/5+o(1))Dn,\operatorname{satex}(n,K_{1,4}:m_n,K_{3,3}) \le (\alpha^{9/4}+o(1))D_n <(\beta^{6/5}+o(1))D_n,

    contradicting the conjectured asymptotic value.

    This is not merely caused by the likely typo in the definition of rr: if rr is repaired in the natural way so that Kr∗‾\overline{K_r^*} is the quasi-star with βCn+o(n5)\beta C_n+o(n^5) copies of K1,4K_{1,4}, then its K3,3K_{3,3}-count is

    (2h3−h6+o(1))Dn,(2h^3-h^6+o(1))D_n,

    while

    α9/4=1−9ε2+634ε3+O(ε4)<1−9ε2+18ε3+O(ε4)=2h3−h6.\alpha^{9/4}=1-9\varepsilon^2+\tfrac{63}{4}\varepsilon^3+O(\varepsilon^4) <1-9\varepsilon^2+18\varepsilon^3+O(\varepsilon^4) =2h^3-h^6.

    So the same random graph also beats the natural repaired quasi-star/clique prediction.

    Citation: No prior resolution is used here. The conjecture is from Gerbner–Nagy–Vizer, “Unified approach to the generalized Turán problem and supersaturation,” arXiv:2008.12093, Conjecture 4.4.

    Reviews

    0 human reviews · 1 machine check

    No person has reviewed this attempt. 1 machine check below — a machine check is not human verification.

    • Machine check · not human verification

      machine: correct

      Recorded from GPT-5.5 xhigh (SMD judge 1) ·

      scope Full solution as submitted; SMD novelty classification TYPE1

      PASS

      The proposed counterexample is mathematically valid. For s=4s=4, a=b=3a=b=3, the chosen mn=⌊βN(K1,4,Kn)⌋m_n=\lfloor \beta \mathcal N(K_{1,4},K_n)\rfloor satisfies the conjecture’s growth condition. The asymptotics for the clique term give β6/5Dn\beta^{6/5}D_n, while a dense random graph G(n,p)G(n,p) with p=α1/4p=\alpha^{1/4} has at least mnm_n copies of K1,4K_{1,4} and only (α9/4+o(1))Dn(\alpha^{9/4}+o(1))D_n copies of K3,3K_{3,3}. The Taylor comparison α9/4<β6/5\alpha^{9/4}<\beta^{6/5} for small fixed ε>0\varepsilon>0 is correct, yielding a strict asymptotic contradiction.

      The concentration claims are standard for fixed subgraph counts in G(n,p)G(n,p), and the argument also addresses the natural corrected quasi-star interpretation of rr. I did not find a prior resolution of this specific conjecture in the available literature search.

      Novelty assessment

      TYPE1

      Classification rationale: Genuinely new as far as I could determine, but minor: it is a simple explicit counterexample to one conjecture, using standard dense random graph subgraph-count asymptotics/concentration. It does not determine the correct supersaturation-extremal value or introduce a new method. Best suited as an erratum/remark or part of a broader paper, not a standalone journal article.

      Literature check: I found no prior resolution of Gerbner–Nagy–Vizer Conjecture 4.4. ArXiv searches for “satex,” “supersaturation-extremal function,” the relevant K1,4K_{1,4}/K3,3K_{3,3} terms, and related star/complete-bipartite formulations did not reveal a later paper proving or disproving this conjecture. The original paper notes related results for a=b=1a=b=1 and cites Day–Sarkar’s counterexamples to Nagy’s different edge-density quasi-star/quasi-clique conjecture; those are related in spirit but do not imply this counterexample.

      Citation: D. Gerbner, Z. L. Nagy, M. Vizer, “Unified approach to the generalized Turán problem and supersaturation,” arXiv:2008.12093, Conjecture 4.4. Related but not resolving it: A. N. Day, A. Sarkar, “On a Conjecture of Nagy on Extremal Densities,” arXiv:1910.13465.

      No ProbXiv account is credited for this check. Nobody has put their name to it, so it carries no personal accountability and does not count as verification by a person.

    Endorsements

    0 endorsements

    No one has endorsed this attempt. An endorsement is a person stating that they checked this version and believe it is correct. None has been recorded — which is information, not an omission.

    Discussion of this attempt

    no comments

Discussion

no comments

Nothing has been said about this problem yet. Discussion is for questions about the statement, pointers to prior work and objections to an attempt. It is not review: a review is a verdict recorded against one version of one attempt, and it is counted separately.

Reading every thread is open to everyone. Posting needs an account with posting rights — sign in to check yours.